Tải bản đầy đủ - 0trang
CHAPTER 9. CYTOGENETICS OF PEARL MILLET
PREM P. JAUHAR
E. Identification of Translocated Chromosomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
F. Disjunction of Interchange Configurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
G. Interchange Heterozygosity and Plant Breeding . . . . . . . . . . . . . . . .
IX. B Chromosomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A. B Chromosomes as Indicators of the Origin of Pearl Millet. . . . . . . . . . . . . . . . . . . 447
B . Mode of Pollination.. . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . 454
XI. Hybridization and Chromosome Relationships . . . . . . . . . . . . . . . . .
A. lntraspecific Hybrids
B. Interspecific Hybrids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C. Intergeneric Hybridization . . . . . . . . . . . . . . . . . . . . . . . . . . . .
XII. Conclusion .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Pennisetum is one of the most important genera of the tribe Paniceae of the
grass family. Pearl millet [Pennisetum ryphoides (Burm.) Stapf et Hubb.] is the
most important constituent of this genus. It is a dual-purpose crop: its grain is
used for human consumption and its fodder serves as feed for cattle. In Asia and
Africa, however, it is grown primarily as a grain crop on an estimated 60 million
acres of relatively poor land. It has remarkable ability to grow in areas of low
rainfall. In sub-Saharan Africa harvests of pearl millet are obtained with as little
as 250 mm of annual rainfall (Brunken, 1977). Its grain is traditionally considered to be nutritious and is put to a variety of uses. As poor man’s bread, it
sustains a large proportion of the populace of Africa and Asia. It also contributes
to the economy of countries like the United States, where it is grown as a forage
crop on an estimated 1 million acres. Pearl millet is also grown as a forage crop
in the tropical and warm-temperate regions of Australia and several other countries.
Pearl millet originated in West Africa. Selection exercised by the early cultivators under a myriad of cultural contexts led to the development of several
morphologically diverse forms. Its protogynous nature facilitated the introgression of characters from other wild and cultivated annual species of the section
Penicillaria. It is now widely cultivated in different parts of the world. In terms of
annual production, pearl millet is the sixth most important cereal crop in the
world, following wheat, rice, maize, barley, and sorghum. Among the millets it
is second only to sorghum. In India, it is the fourth most important cereal after
rice, wheat, and sorghum.
CYTOGENETICS OF PEARL MILLET
Pearl millet is a favorable organism for genetic research. Its chromosome
number, 2n = 14, was determined more than 50 years ago by Rau (1929).
Several favorable features of the chromosome complement, e.g., small number
and large size of chromosomes with one distinctive pair of nucleolar organizers,
make pearl millet a suitable organism for cytogenetic studies. Moreover, its
protogynous flowers and outbreeding system make it ideal for interspecific hybridization and for breeding work. It is indeed ideally suited for heterosis breeding. Although pearl millet has a remarkable ability to grow on soils of marginal
fertility, it responds very well to proper fertilization, which helps in realizing the
high yield potential of its hybrids. The hybrids’ greater N use efficiency (biomass
production per unit of N in the plant) is probably attributable to the highly
efficient (C,) photosynthetic pathway of this crop.
Although pearl millet has great agricultural importance, is a very favorable
organism for cytogenetic studies and breeding work, and has a low chromosome
number that was also determined at about the same time as those of most other
crops, the information available on its genetics and cytogenetics is much less
than that known for other important crops. There are several reasons why this
crop has been largely overlooked as a genetic and cytogenetic tool:
1. It has long been considered to be a crop of secondary importance and, thus,
could not compete for research funding with other crops like wheat and corn.
2 . It has a restricted area of use, being a food for the poor only, although it is
also an excellent fodder crop.
3. Its potential as a research tool was not appreciated until recently.
4. The existence of long-standing nomenclatural controversies (in the postLinnaean period from 1753 to 1759, pearl millet has been treated as a member of
at least six different genera, viz., Panicum, Holcus, Alopecurus, Cenchrus,
Penicillaria, and Pennisetum, and has been given different botanical names; see
Jauhar, 1981a) could also have had an adverse impact on research.
Studies on chromosome pairing in interspecific hybrids-with pearl millet as
one of the parents-have contributed to our understanding of phylogenetic relationships between different P ennisetum species and pearl millet. In these studies,
the large size of the pearl millet chromosomes has been helpful in ascertaining
chromosome relationships. Because of its low chromosome number, pearl millet
also offers a particularly favorable material for aneuploid analyses, which should
be helpful in the elucidation of its cytogenetic architecture. Primary trisomics
constitute a valuable tool for locating genes on different chromosomes and for
assigning them to linkage groups. Although considerable progress has been made
in developing a set of primary trisomics in pearl millet, the establishment of
linkage groups awaits completion. A good deal of information is available on
certain other cytogenetic aspects, e.g., polyploidy, interchange heterozygosity ,
haploidy, and B chromosomes. All these studies should contribute to the improvement programs of pearl millet.
PREM P. JAUHAR
The purpose of this article is to summarize and integrate the available information on different aspects of pearl millet cytogenetics. It is hoped that this article
will provide useful information to cytogeneticists and breeders engaged in the
improvement of pearl millet and other forage species of Penniseturn. This article
may also be of interest to a spectrum of other workers engaged in basic research.
II. KARYOTYPIC ANALYSIS
Karyotypic analysis includes the study of the number, size, and morphology of
chromosomes. Total length and arm ratios of chymosomes are helpful in systematic and phylogenetic investigations. Levitskii (193 1) and Avdulov (1 93 1)
pioneered the use of cytological features as aids in establishing taxonomic and
phylogenetic relationships among species and genera. Although basic number,
size, and morphology of the chromosomes can indeed be useful in taxonomic
classification (Hunter, 1934; Constance, 1957), these parameters should be subsidiary to morphological characters in any taxonomic treatment (Pilger, 1954).
Modern cytological techniques, e.g., the banding of chromosomes with Giemsa
(Vosa and Marchi, 1972; Vosa, 1973, 1975), and staining heterochromatic patterns with fluorochromes like quinacrine mustard (Vosa, 1970) can provide information of phylogenetic value.
The occurrence of cytotypes or chromosomal races (intraspecific polyploid
series) is a characteristic feature of the perennial species of Penniseturn. However, no such cytotypes exist in the annual cultivated or wild pearl millets, which
all have 2 n = 14 chromosomes (Table I); in fact, all these taxa belong to the
species P . ryphoides. There is a report of 2 n = 36 chromosomes for a Nigerian
collection of “ P . violaceurn (Lam.) L . Rich.” (Olorode, 1975), but this could be
an incorrect identification. Since the material classified as P . violaceurn forms
fully fertile hybrids with pearl millet (2n = 14), the former must have 2 n = 14
chromosomes (see Section XI,A).
Chromosomes are generally measured at somatic metaphase after pretreatments that condense and spread them. The main drawback inherent in these
studies is that the magnitude of error in the measurements of condensed chromosomes is high. Therefore relatively small size differences among chromosomes
of a species, of infraspecific categories, or of different species cannot be resolved
accurately. However, karyomorphological studies can be done more precisely on
pachytene chromosomes in taxa with low chromosome numbers, e.g., P .
CYTOGENETICS OF PEARL MILLET
Chromosome Numbers of Different Taxa in the Section Penicillaria of Genus Penniserurn
Cultivated pearl millet
P. typhoides (Burm.) Stapf et Hubb.
[Syn. P. tvphoideutn Rich.
P. spicarurn (L.) Koern.
P. glaucum (L.) R. Br.
P. amerironum ( L . ) K. Schum.]
Annual relatives of pearl millet"
P. alhicauda Stapf et Hubb.
P. rmcvhchaele Stapf et Hubb.
P. rinereum Stapf et Hubb.
P. dalzielii Stapf et Hubb.
P. echinurus (K. Schum.) Stapf et Hubb.
P. fullax (Fig. & De Not.) Stapf et Hubb.
P. gambiense Stapf et Hubb.
P. leanis Stapf et Hubb.
P. tnuiwa Stapf et Hubb.
P. nigrilarurn Schlecht.
P. mollissimum Hochst.
P. perrottetii (Klotzch ex A.Br.)K. Schum
P. pyrnostachyum (Steud.) Stapf et Hubb.
P. pynostachvunr var. gambia
P. versicolor Schrad.
P. violareum (Lam.) L. Rich
Perennial relative of pearl millet
P. purpureutn Schum.
Krishnaswamy (1951);Thevenin (1952)
Bilquez and Lecomte (1969)
Mehra et ul. (1968)
Burton (1942); Nishiyama and Kondo
Krishnaswamy and Raman (1948);
Gadella and Kliphuis (1964)
"These and other annual, cultivated, or wild relatives of pearl millet have 2n = 14 chromosomes.
They are not reproductively isolated from the cultivated species-P. fyphoides-and in fact do not
deserve specific ranks.
ramosum ( 2 n = 10) and P . typhoides ( 2 n = 14). For critical comparisons, the
DNA content of chromosomes can also be measured.
The genus Pennisetum is a heterogeneous assemblage of species with chromosome numbers ranging from 2 n = 10 to 2 n = 7 2 , being multiples of 5, 7, 8, and
9. Their chromosome morphology is also very diverse, with tremendous size
differences; a noteworthy feature is that the species with lower numbers have the
larger sizes. Thus, pearl millet ( P . ryphoides) has only 2 n = 14, but relatively
very large chromosomes. Avdulov (1931) noted that pearl millet had 14 large
chromosomes, larger than those of any other member of the tribe Paniceae.
PREM P. JAUHAR
However, I think that the annual (or rarely biennial) species P . ramosum (2n =
10) has the largest chromosomes in the genus Pennisetum and probably in the
entire tribe Paniceae. The chromosomes of P . ramosum are approximately 5%
larger than those of P . typhoides. Thus, in the genus Pennisetum, the species
with the lowest chromosome number (2n = 10) has the largest chromosomes.
In contrast, the species with higher chromosome numbers (e.g., P . orientale,
2n = 18, 36, 54) have strikingly smaller chromosomes than those of P .
ramosum or P . typhoides. The trend of species with low chromosome numbers to
have much larger chromosomes is evident in several other plant groups. In
Sorghum, for example, the average lengths of chromosomes of S. versicolor (2n
= lo), S . vulgare (2n = 20), and S . halepense (2n = 40) were 4.86, 2.24, and
1.98 p m , respectively (Karper and Chisholm, 1936).
1 . P . typhoides (2n = 1 4 )
Rau (1929) determined from root tips the chromosome number of pearl millet
as 2n = 14. Moreover, he mentioned that “the chromosomes are very large” and
that the homologous pairs could be easily distinguished. Avdulov (1931) studied
the chromosomes of pearl millet, which was at that time classified as Penicillaria
spicata Willd. His drawing shows 14 chromosomes with median to submedian
centromeres, the shortest chromosome being the satellited one. It is interesting to
note that as early as 1931 when cytological techniques were not perfected,
Avdulov noticed one pair of satellited chromosomes; this observation has been
confirmed by numerous workers. The small nucleolar bivalent is clearly observed to be associated with the nucleolus (Fig. 1 ) .
Pantulu (1 958) examined the chromosomes at pachytene and grouped them
into four classes on the basis of relative length and position of centromere: ( 1 )
two large pairs (chromosomes 1 and 2) with median centromeres; ( 2 ) two somewhat shorter pairs (chromosomes 3 and 4) with median to submedian centromeres; (3) two medium-sized pairs (chromosomes 5 and 6) with submedian
centromeres; and (4)the shortest pair (chromosome 7) with the nucleolus organizer. Later, essentially similar results were obtained on the analysis of
karyotype at pachytene (Venkateswarlu and Pantulu, 1968; and Lobana and Gill,
1973), at pollen mitosis (Krishnaswamy and Raman, 1953a), and at somatic
metaphase (Burton and Powell, 1968). Virmani and Gill (1972) and Tyagi
(1975a) karyotyped the somatic chromosomes and classified them as follows:
chromosomes 1 , 2, 3, and 5 as metacentric; chromosomes 4 and 6 as submetacentric; and chromosome 7 as subterminal.
Thus, there are minor disagreements among different workers as to the position of centromere. Looking at a condensed chromosome at somatic metaphase,
it is not unexpected that one worker locates the centromere as median, whereas
another classifies it as submedian. The same workers, looking at pachytene and
somatic chromosomes, can also arrive at different conclusions. For example,
CYTOGENETICS OF PEARL MILLET
FIG. 1. Diakinesis in pearl millet ( 2 n = 14) showing one nucleolar bivalent; note small rod
associated with the nucleolus. Also note chiasma terminahation. [ X 12601
Virmani and Gill (1972) studied somatic chromosomes and classified chromosome 1 as metacentric; whereas, based on pachytene analysis, Lobana and Gill
(1973) considered it to be submetacentric.
There is no doubt that pearl millet has a fairly symmetrical karyotype. It is
certainly not very easy to identify all of the seven chromosomes by the techniques
currently used; therefore, Giemsa banding (see Vosa, 1973, 1975; Zelleret al.,
1977; Filion and Blakey, 1979) of somatic prometaphase chromosomes must be
tried to identify individual members of the complement. It has mostly metacentric or submetacentric chromosomes, the longest being approximately 1.5
times the shortest; both these features are indices of symmetry of karyotype.
Under Stebbins’ (1958) classification of types of asymmetry, pearl millet will fit
best in the class la, i.e., the most symmetrical of the 12 karyotypes described.
The shortest chromosome pair is somewhat subterminal with the satellite on its
short arm. It can be identified in somatic plates as well as at pachytene and
diakinesis, where, as a small bivalent, it is associated with the nucleolus (Fig. 1).
Chromosomes of some diploid taxa of the section Penicillaria, which are
annual relatives of pearl millet, have been observed. The materials classified as
Pennisetum cinereum, P. echinurus, P . gumbiense, P. leonis, and P. pycnosruchyum had 2n = 14 chromosomes, as in cultivated pearl millet (Krishnaswamy, 1951). Veyret (1957) found that P. ancylochaete, P . gambiense, P .
maiwa, and P. nigritarum had 2n = 14 chromosomes, and their chromosome
morphologies were similar to one another and also to that of cultivated pearl
millet. Genetic studies by Bilquez and Lecomte (1969) and Brunken (1977)
PREM P. JAUHAR
have shown that P . violuceum and P . fullux-two of the important wild, annual
relatives of pearl millet-are not reproductively isolated from it; their hybrids
with pearl millet were highly fertile. Although these workers have not mentioned
the chromosome number of these wild taxa, they obviously have 2n = 14
chromosomes in order to form fertile hybrids with pearl millet.
2. P . purpureum (2n
= 4x =
Napier grass, an allotetraploid and a relative of pearl millet, has a somewhat
asymmetrical karyotype consisting of chromosomes with median, submedian,
and subterminal centromeres. On the basis of pachytene studies, Pantulu and
Venkateswarlu (1968) reported that the longest chromosome of the complement
(chromosome 1) was 2.7 times the length of the shortest (chromosome 14), thus
making the karyotype asymmetrical. Based on these observations, the karyotype
of P . purpureum will fall in the category 2b in Stebbins’ (1958) classification of
types of asymmetry.
Pantulu and Venkateswarlu (1968) reported that chromosomes 1 and 14 have
nucleolus organizers. The largest chromosome of the complement (chromosome
1) is certainly satellited, as evidenced by the association of the largest bivalent
with the nucleolus (Fig. 2 ) . The other nucleolar bivalent is one of the smallest, if
not the smallest, in the complement. If chromosome 14 is indeed satellited, then
FIG.2. Diakinesis in napier grass ( 2 n = 4x = 28) showing 14 bivalents (9 rings and 5 rods) with
terminalized chiasmata. Note one large bivalent (the largest in the complement) and one relatively
small bivalent associated with the nucleolus. Also note an additional small nucleolus (marked with
arrow). [ x 12601
CYTOGENETICS OF PEARL MILLET
P. purpureum shares an important karyotypic feature with P. typhoides, i.e., the
shortest chromosomes of both the species are satellited. Moreover, during meiotic prophase both typhoides and purpureurn show rapid terminalization of chiasmata (see Section 111). They also seem to have similar patterns of centromeric
heterochromatin. Thus, P. typhoides and P. purpureum seem to share some
important karyological features of phyletic value.
All penicillarias fall into the x = 7 group (see Table I). They have conspicuously penicillate anther tips (see Fig. 14a,b). Of these, only one species ( P .
purpureum) is a perennial tetraploid. All other taxa are annual and diploid with
2n = 2x = 14 chromosomes. The annual, semiwild taxa are not reproductively
isolated from the cultivated pearl millet and must be considered as infraspecific
categories within P . typhoides. They have regular meiosis with 7,,, as in P.
A . P . ryphoides ( B u R M . )STAPFET HUBB.( 2 n = 2x = 14)
Rau (1929) determined the chromosome number of pearl millet as 2 n = 14.
Rangaswamy (1935) studied meiosis and found at diakinesis mostly seven ringshaped bivalents having two terminalized chiasmata each.
In different populations of pearl millet, mostly ring bivalents with two chiasmata each are observed at diakinesis, but the nucleolar bivalent is generally a
small rod with one chiasma (Figs. 1 and 3b). The rapid terminalization of
chiasmata seems to be a characteristic feature, so that at diakinesis the bivalents
generally appear loose and dissociated (Figs. 1 and 3b,c). At metaphase, both
ring and rod bivalents are observed. In some cultivated varieties in India, the
mean chiasma frequency at metaphase was found to be 12.10 per cell and 0.86
per paired chromosome; this means that ring bivalents are preponderant (see Fig.
Some populations of pearl millet show secondary associations of bivalents.
Two groups of two bivalents each were clearly observed (Fig. 3c) in some cells.
Although the phyletic significance of secondary associations in diploid species
remains controversial, such associations cannot be entirely meaningless. In
hexaploid wheat, such associations are known to take place between genetically
and evolutionarily related chromosomes (Riley, 1960; Kempanna and Riley,
1964). In pearl millet, the secondarily associated bivalents look very similar to
each other, although their genetic and phyletic relatedness cannot be determined.
In the haploid complement when their homologous partners are missing, the
chromosomes involved in these secondary associations probably form bivalents.
PREM P. JAUHAR
FIG. 3. Meiotic stages in pearl millet. (a) Late diplotene showing 7 bivalents (711).(b) Diakinesis
with 711with teminalized chiasmata. Note 6 ring bivalents and the small, nucleolar rod bivalent. (c)
Diakinesis with 711.Note the secondary associations of two pairs of bivalents; the associated bivalents
look similar in size and shape. (d) Metaphase 1 with 611.[(a, d) X ca. 2050; (b, c) x ca. 21501
CYTOGENETICS OF PEARL MILLET
It is interesting to note that two bivalents have been reported in haploids studied
by different workers (see Section V,B, Table 11; Fig. 6c). These observations
lend favor to the suggestion that the complement of typhoides has been derived
from a basic set of x = 5 chromosomes (see Jauhar, 1968, 1970b; Sections V,B
B . P . purpureum SCHUMACH.
( 2 n = 28, 56)
Elephant or napier grass ( P . purpureum) is a perennial relative of pearl millet
and is native to Africa. Burton (1942) and Nishiyama and Kondo (1942) determined its somatic chromosome number as 2n = 28, which is tetraploid based on
x = 7. It shows diploid-like meiosis, 14,, being regularly formed at diplotene,
diakinesis, and metaphase (Fig. 4a-c). No multivalents or univalents are generally formed. The occasional occurrence of a quadrivalent (Olorode, 1974) can be
attributed to a floating interchange in certain populations.
At diakinesis, there is a rapid terminalization of chiasmata (Figs. 2 and 4b)-a
feature also characteristic of pearl millet. Two bivalents are generally associated with the nucleolus (Fig. 4a,b). One of the nucleolar bivalents is the
largest in the complement, whereas the other is a small one (see also Section
11,2). Occasionally, additional nucleolar material is organized (see Fig. 2). At
metaphase, there are noticeable size differences among bivalents; the smaller
ones are generally rod-shaped with one chiasma each, whereas the majority of
the large ones are ring-shaped with mostly two chiasmata each (Fig. 4c).
Chiasma frequency per cell and per paired chromosome was found to be 18.9
and 0.68, respectively, in some collections.
Several factors speak for the allotetraploid nature of elephant grass: ( I ) its 2n
= 28 chromosome number; (2) the regular bivalent formation and high pollen
fertility; and (3) the noticeable size differences among bivalents. This is further
borne out by studies on chromosome pairing in its hybrids with pearl millet (see
Section XI,B,2,d). This allotetraploid can be genomically represented as
A'A'BB. The A' genome is homoeologous with the A genome of pearl millet
(which is genomically AA). The large bivalents observed at metaphase in P .
purpureum are evidently formed by the A'A' genome, whereas the small ones
belong to the BB genome, the donor of which is not yet known.
IV. ABNORMAL MEIOSIS AND ITS GENETICS
The nature of events that lead to synapsis and crossing over during the meiotic
prophase remains one of the most intriguing problems in cytogenetics today. It is
PREM P. JAUHAR
FIG. 4. Meiotic stages in napier grass ( 2 n = 4x = 28). (a) Early diakinesis with 14 bivalents
(l4,,). Note that two bivalents (one the largest in the complement and one much smaller) are
associated with the nucleolus. (b) Diakinesis with 14,: Due to terminalization of chiasmata, most
bivalents appear to be loose and dissociated. Note that 2,, are associated with the nucleolus; the large
bivalent-largest in the complement-is lying on the nucleolus. (c) Metaphase I with 14,,; I,, is
separated. [ x 14801